PHYLOGENY. 



157 



who adopts (without credit) my hypothesis of lemurine 

 affinities of the Condylarthra (which he renames the 

 Mesodactyla). From Condylarthra back to Creodonta 

 is an easy transition, and I have always assumed that 

 the Creodonta were derived from generalized polypro- 



todont Mursupialia. This view 

 has been entirely confirmed by 

 the recent discoveries of Ame- 

 ghino in Patagonia, where he 

 has found forms whose remains 

 may be referred with equal pro- 

 priety to the one group or the 

 other. M. Topinard has been 

 rather hasty in reaching the 

 marsupial ancestry in suppos- 

 ing that Phenacodus belongs to 

 that order. All the evidence 

 shows that Phenacodus is a 

 generalized ungulate placental. 

 To return to the more im- 

 mediate ancestry of man. I 

 have expressed, 1 and now main- 

 tain as a working hypothesis, 

 that all the Anthropomorpha 

 were descended from the Eo 

 cene lemuroids. In my sys- 

 tem 2 the Anthropomorpha in- 

 cludes the two families Homi- 

 nidae and Simiidae. The sole 

 difference between these families is seen in the struc- 

 ture of the posterior foot ; the Simiidae having the 



^American Naturalist, 1885, p. 467. 



2 Origin of the Fittest, 1887, p. 346, from American Naturalist, 1885, p. 34*, 

 where the classification of the Taxeopoda should be in a foot-note; loc. cit., 

 1889, October. 



Fig. 40. Tomitherium ros- 

 tratum Cope, fore arm, five- 

 sixths natural size. Original, 

 b, ulna; c, radius. 



